NASA Technical Reports Server (NTRS) 20050160169: Chirality Characterization of Dispersed Single Wall Carbon Nanotubes

0.14 MB·English

by  NASA Technical Reports Server (NTRS)

Chirality Characterization of Dispersed Single Wall Carbon Nanotubes

Min Namkung1, Phillip A. Williams2, Candis D. Mayweather3, Buzz Wincheski1, Cheol Park4,
and Juock S. Namkung5
1NASA Langley Research Center, Hampton, VA 23681
2National Research Council, NASA Langley Research Center, Hampton, VA 23681
3Spelman College, Atlanta, Georgia 30314
4National Institute of Aerospace, Hampton, VA 23666
5Naval Air Warfare Center, Patuxent River, MD 20670

ABSTRACT

Raman scattering and optical absorption spectroscopy are used for the chirality characteriza-
tion of HiPco single wall carbon nanotubes (SWNTs) dispersed in aqueous solution with the
surfactant sodium dodecylbenzene sulfonate. Radial breathing mode (RBM) Raman peaks for
semiconducting and metallic SWNTs are identified by directly comparing the Raman spectra
with the Kataura plot. The SWNT diameters are calculated from these resonant peak positions.
Next, a list of (n, m) pairs, yielding the SWNT diameters within a few percent of that obtained
from each resonant peak position, is established. The interband transition energies for the list of
SWNT (n, m) pairs are calculated based on the tight binding energy expression for each list of
the (n, m) pairs, and the pairs yielding the closest values to the corresponding experimental
optical absorption peaks are selected. The results reveal that (1, 11), (4, 11), and (0, 11) as the
most probable chiralities of the semiconducting nanotubes. The results also reveal that (4, 16),
(6, 12) and (8, 8) are the most probable chiralities for the metallic nanotubes. Directly relating
the Raman scattering data to the optical absorption spectra, the present method is considered the
simplest technique currently available. Another advantage of this technique is the use of the ES
11
peaks in the optical absorption spectrum in the analysis to enhance the accuracy in the results.

INTRODUCTION

The physical properties of a single wall carbon nanotube (SWNT) depend on its chirality,
which is determined by two integer coefficients, normally denoted by (n, m) [1]. The spectros-
copy methods of fluorescence, Raman scattering and optical absorption have been the
mainstream tools for chirality characterization [2-4], and each method has unique capabilities.
Fluorescence spectroscopy is a novel technique providing information on the chirality of SWNTs
but its applicability is limited to semiconducting SWNTs. The Raman scattering technique
directly provides the distribution of SWNT diameters which are closely related to the chirality as
the diameter of an individual SWNT is determined by its (n, m). Optical absorption spectros-
copy provides the interband transition energies, which are also closely related to the chirality
distribution since the interband transition energies of a SWNT are an explicit function of its (n,
m) values. Unfortunately, both the SWNT diameter and interband transition energy are multi-
valued functions of (n, m). Hence, it is logical to combine the results of the Raman and optical
absorption spectra of a sample for a unique determination of the chirality distribution. However,
the wavelength of the excitation radiation for the optical absorption method varies continuously
within a range, whereas the monochromatic laser beams used for the Raman spectroscopy
possibly leave unmatched spectral regions.
Recently, two research groups in the US and Europe have employed tunable laser sources
and obtained Raman spectra with the incident laser wavelength varying in a range [3,5,6]. This
technique provides not only the Raman spectra but also the optical absorption spectra from a
batch of SWNTs at the same time. This elegant technique ensures the one-to-one correspond-
dence between specific resonant Raman peaks and optical absorption peaks enabling a complete
incorporation of two spectra in a given range. Despite such a clear advantage, the experimental
procedure of this scanned Raman method with tunable lasers can be complex such that its
application may be limited to research purposes only. In this paper, we present a simpler
procedure that is capable of characterizing the chirality distribution of a batch of SWNTs. The
method is based on the direct combination of the Raman and optical absorption spectra. In
particular, we demonstrate that a near complete comparison between the two types of spectra is
accomplished with limited Raman data.

EXPERIMENTAL DETAILS

Purified SWNTs synthesized through the HiPco process [7] are obtained from Carbon
Nanotechnologies, Inc. Following procedures similar to those of Islam et al. [8] and Paredes et
al. [9], the SWNT material is dispersed in an aqueous solution of the surfactant sodium dodecyl
benzene sulfonate (NaDDBS). Briefly, the SWNT material as dispersed in the surfactant
solution is subjected to ultrasonic agitation and centrifugation followed by decanting of the
supernatant. This process suspends the SWNTs in the solution and exfoliates the larger bundles
of SWNTs into smaller bundles and individual nanotubes for more accurate quantitative analysis
of the SWNT chirality distribution.
Optical absorption measurements of the resulting supernatant are taken with a Perkin Elmer
UV-vis spectrometer with a slit size of 5 nm and range of 200-1300 nm. Dispersive Raman
spectra (Kaiser Optical Systems, Inc 5000R system) are acquired with excitation wavelengths of
632.8 nm and 785 nm, and Fourer Transform (FT)-Raman spectra (Raman accessory to Digilab
7000 FTS) are also acquired with a laser excitation wavelength of 1064 nm. Baseline correction
and peak fitting analysis are applied on all acquired spectra for the determination of peak
positions.

RESULTS AND DISCUSSION

Figure 1 shows the Raman radial breathing mode (RBM) peaks of a spectrum from a sample
of the HiPco SWNTs dispersed in an aqueous solution of NaDDBS obtained with the incident
laser wavelength of 632.8 nm. As indicated, seven resonant peaks are observed at ω =
RBM
197.79, 222.20, 238.53, 254.31, 265.07, 284.49, and 297.18 cm−1. The dispersive Raman data
acquired with an excitation wavelength of 785 nm exhibit four RBM peaks at ω = 172.25,
RBM
207.62, 229.59, and 266.73 cm−1. For the FT-Raman spectra acquired with an excitation
wavelength of 1064 nm, a single RBM peak is present atω =267.56cm−1. The SWNT
RBM
diameters, d , are calculated from the RBM peak positions using the expression
t

ω =239(cm−1nm)/d (nm)+8.5cm−1 (1)
RBM t
2000
-1
254.31 cm
-1
1800 284.49 cm
1600
-1
197.79 cm
-1
1400 265.07 cm
-1
238.53 cm
1200
y
t
i
s -1
n 222.20 cm
e 1000
nt 297.18 cm-1
I
800
600
400
200
0
150 170 190 210 230 250 270 290 310
-1
Raman Shift (cm )

Figure 1. Radial breathing mode (RBM) Raman scattering peaks from SWNTs obtained with an
incident laser excitation wavelength of 632.8 nm. The solid trace is the acquired data, and the
dotted lines indicate resulting peaks fitted through spectral analysis.

which has been found experimentally to be most accurate for isolated HiPco SWNTs [10]. A list
of (n, m) pairs producing the d values within 10% of each obtained from a RBM peak position
t
is then constructed. The interband transition energy for each (n, m) pair is then calculated using
a first nearest neighbor tight binding energy expression [11]

r r r
r r r r
E(k) = ±γ {3−cosk ⋅a −cosk ⋅a −cosk ⋅(a −a ) (2)
o 1 2 1 2
+ 3sinkr⋅ar − 3sinkr⋅ar − 3sinkr⋅(ar −ar )}1/2
1 2 1 2

r
where γ = 2.9 eV is the nearest neighbor overlap energy, k is a wave vector originated from a
o
Fermi level that is normally denoted as the K-point in a two-dimensional hexagonal reciprocal
r r
lattice, and a and a are the primitive lattice translation vectors of a graphite sheet. The actual
1 2
r r r r r
calculation usesk = k +k , wherek andk are the components which are parallel and
z θ z θ
perpendicular to the SWNT axis, respectively. The (n, m) dependence of E(k) originates from
r r
that in the expressions of the chiral angle and discrete k [11]. At a given k the magnitude of
θ θ
r
E(k) is numerically calculated at k where dE/dk | = 0. To include the density of
z,o z k =k
z z,o
r
states (DOS) peak split effect [11,12], the calculation is repeated for -k , and the average of
z,o
the two is doubled to compute the interband transition energy. The interband transition energies
thus calculated are compared with the peak positions in the optical absorption spectrum shown in
Figure 2, and the results are summarized in Table I.
As shown in Table I, the chiralities present in the batch of SWNTs from which the surfactant
suspensions were made include (4, 16), (6, 12), and (8, 8) among the metallic nanotubes. The
data also show matches for the chirality assignments of (4, 11), (1, 11), (2, 12), (5, 9), and (0, 11)
for the semiconducting SWNTs. As highlighted in Table I, all three excitation wavelengths
correlate to within a few percent for the chirality assignment of (1, 11) based on the first and
second semiconducting interband transitions and the calculated diameter.
This method and resulting data demonstrate that a significant level of (n, m) characterization
is possible with limited Raman data when these are directly incorporated into the optical
absorption spectrum. The batch of HiPco SWNTs studied here has a rather broad diameter
distribution such that the ES and EM optical absorption peaks overlap, as can be seen in Figure
22 11
2. If the low energy limit of the optical absorption spectroscopy can be lowered, one can
experimentally observe a broader range of the ES and ES absorption peaks from the same
11 22
SWNTs for more extensive comparison with the theoretical transition energies. The same is true

0.5 0.5
S M M
11 11 22
0.4 0.4
S
22
)
.
u
a.0.3 0.3
(

e
c
n
a
b
r
so0.2 + * 0.2
b
A
0.1 0.1
0 0
0.5 1.5 2.5 3.5
Energy (eV)

Figure 2. Optical absorption spectrum of the SWNT sample. The dotted line trace denoted by *
(on the right in the figure) is the acquired, unprocessed data, and the solid line trace denoted by +
is the spectrum resulting from the fitted peaks (also shown with solid line traces). The range of
the interband transition energies for metallic (Mii) or semiconducting (Sii) nanotubes covered by
the RBM peaks of the experimental Raman spectra are indicated by the dashed vertical lines.
Table I. Chirality assignment of SWNTs from Raman and optical absorption data.
λ ω d d Transition ∆E ∆E
excit RBM calculated theory theory opt abs
Peak (nm) (cm-1) (nm) (nm) (n, m) (eV) (eV)
1 785 172.25 1.4596 1.4528 4, 16 M11 1.70814 1.6998
1.4596 1.4528 4, 16 M22 3.3245 3.3338
2 785 207.62 1.2003 1.1994 5, 12 S22 1.3806 1.4306
3 785 229.59 1.081 1.066 4, 11 S22 1.5513 1.5522
4 785 266.73 0.9255 0.9141 1, 11 S22 1.8133 1.8368
0.9255 0.9141 1, 11 S11 0.911 0.9368
5 1064 267.56 0.9226 0.9141 1, 11 S11 0.911 0.9368
0.9226 0.9141 1, 11 S22 1.8133 1.8368
6 632.8 265.07 0.9315 0.9141 1, 11 S22 1.8133 1.8368
0.9315 0.9141 1, 11 S11 0.911 0.9368
7 632.8 197.79 1.2623 1.2582 6, 12 M11 1.9562 1.9104
8 632.8 222.2 1.1184 1.0983 8, 8 M11 2.2196 2.2313
9 632.8 238.53 1.039 1.039 2, 12 S22 1.5954 1.6287
10 632.8 254.31 0.9723 0.974 5, 9 S22 1.6909 1.6998
11 632.8 284.49 0.86597 0.8719 0, 11 S22 1.9012 1.9104
0.86597 0.8719 0, 11 S11 0.9549 0.9783

for EM and EM if the upper limit of the optical absorption spectral range can be increased. The
11 22
accuracy of the (n, m) assignment depends critically on the accuracy of the tight binding
calculation. Reich et al. demonstrated that there exist practically no discrepancies among ES
11
and ES calculated by the ab initio, nearest neighbor tight binding and third neighbor tight
22
binding methods [13]. However, the nearest neighbor tight binding method overestimates the
values of EM by approximately 10% as compared to those computed by the other two methods,
11
and the discrepancy is much more pronounced in the values of EM. Thus, our current
22
assignment of (n, m) values for metallic SWNTs may not be totally accurate, necessitating the
use of the third nearest neighbor tight binding method in the future work.

CONCLUSIONS

In this paper, we demonstrate a near complete characterization of chirality of a batch of
SWNTs by directly incorporating the optical absorption spectrum into limited RBM Raman
spectra. The peaks of these two types of spectra almost completely match over the entire
spectral range of the optical absorption spectra. It is also noted that the accuracy of chirality
characterization can be enhanced both for the semiconducting and the metallic SWNTs if the
optical absorption spectral range is extended. Involving only non-specialized tests of Raman and
optical absorption spectroscopy and a simple analysis procedure, the method in the future can be
developed as a practical tool for routine characterization of the chirality distribution of batches of
SWNTs.
ACKNOWLEDGMENTS

The authors thank Dr. Kristin Pawlowski Burney for her assistance in sample preparation.
The research was performed while the author P. Williams held a National Research Council
Research Associateship Award at NASA Langley Research Center and C. Mayweather was a
participant in the NASA Summer Scholars Program.

REFERENCES

1. R. Saito, G. Dresselhaus, and M.S. Dresselhaus, Physical Properties of Carbon Nanotubes.
(Imperial College Press, London, 2001).
2. M.J. O'Connell, S.M. Bachilo, C.B. Huffman, V.C. Moore, M.S. Strano, E.H. Haroz, K.L.
Rialon, P.J. Boul, W.H. Noon, C. Kittrell, J. Ma, R.H. Hauge, R.B. Weisman, and R.E.
Smalley, Science 297 (5581), 593 (2002).
3. M.S. Strano, S.K. Doorn, E.H. Haroz, C. Kittrell, R.H. Hauge, and R.E. Smalley, Nano
Letters 3 (8), 1091 (2003).
4. A. Hartschuh, H.N. Pedrosa, L. Novotny, and T.D. Krauss, Science 301 (5638), 1354 (2003).
5. C. Fantini, A. Jorio, M. Souza, M.S. Strano, M.S. Dresselhaus, and M.A. Pimenta, Phys. Rev.
Lett. 93, 147406 (2004).
6. H. Telg, J. Maultzsch, S. Reich, F. Hennrich, and C. Thomsen, Phys. Rev. Lett. 93 (17),
177401 (2004).
7. P. Nikolaev, M.J. Bronikowski, R.K. Bradley, F. Rohmund, D.T. Colbert, K.A. Smith, and
R.E. Smalley, Chem. Phys. Lett. 313 (1-2), 91 (1999).
8. M.F. Islam, E. Rojas, D.M. Bergey, A.T. Johnson, and A.G. Yodh, Nano Letters 3 (2), 269
(2003).
9. J.I. Paredes and M. Burghard, Langmuir 20 (12), 5149 (2004).
10. A. Kukovecz, C. Kramberger, V. Georgakilas, M. Prato, and H. Kuzmany, European
Physical Journal B 28 (2), 223 (2002).
11. S. Reich and C. Thomsen, Physical Review B (Condensed Matter and Materials Physics) 62
(7), 4273 (2000).
12. R. Saito, G. Dresselhaus, and M.S. Dresselhaus, Physical Review B (Condensed Matter and
Materials Physics) 61 (4), 2981 (2000).
13. S. Reich, C. Thomsen, and P. Ordejon, Physical Review B (Condensed Matter and Materials
Physics) 65, 155411 (2002).